import 'dart:math';
import 'dart:typed_data';
import 'dart:ui';
import 'package:image/image.dart' as img;
import 'package:tflite_flutter/tflite_flutter.dart';
import 'package:cloud_firestore/cloud_firestore.dart';
import 'Recognition.dart';
class Recognizer {
late Interpreter interpreter;
late InterpreterOptions _interpreterOptions;
static const int WIDTH = 112;
static const int HEIGHT = 112;
final FirebaseFirestore firestore = FirebaseFirestore.instance;
Map<String, Recognition> registered = {};
Recognizer({int? numThreads}) {
_interpreterOptions = InterpreterOptions();
if (numThreads != null) {
_interpreterOptions.threads = numThreads;
}
loadModel();
loadRegisteredFaces();
}
Future<void> loadRegisteredFaces() async {
try {
QuerySnapshot snapshot =
await firestore.collection('registered_faces').get();
for (QueryDocumentSnapshot doc in snapshot.docs) {
String name = doc['name'];
List<double> embedding =
(doc['embedding'] as String).split(',').map(double.parse).toList();
registered[name] = Recognition(name, Rect.zero, embedding, 0);
}
print('Loaded registered faces: ${registered.length}');
} catch (e) {
print('Error loading registered faces: $e');
}
}
Future<void> registerFaceInDB(String name, List<double> embedding) async {
try {
// Create a map to store the face information
Map<String, dynamic> faceData = {
'name': name,
'embedding':
embedding.join(","), // Convert embedding to comma-separated string
};
// Add the data to the 'registered_faces' collection in Firebase
await firestore.collection('registered_faces').doc(name).set(faceData);
print('Successfully registered face: $name');
} catch (e) {
// Log an error message if something goes wrong
print('Error registering face: $e');
}
}
Future<void> loadModel() async {
try {
interpreter =
await Interpreter.fromAsset('assets/mobile_face_net.tflite');
} catch (e) {
print('Unable to create interpreter, Caught Exception: ${e.toString()}');
}
}
List<dynamic> imageToArray(img.Image inputImage) {
img.Image resizedImage =
img.copyResize(inputImage, width: WIDTH, height: HEIGHT);
List<double> flattenedList =
resizedImage.getBytes().map((value) => value.toDouble()).toList();
Float32List float32Array = Float32List.fromList(flattenedList);
int channels = 3;
int height = HEIGHT;
int width = WIDTH;
Float32List reshapedArray = Float32List(1 * height * width * channels);
for (int c = 0; c < channels; c++) {
for (int h = 0; h < height; h++) {
for (int w = 0; w < width; w++) {
int index = c * height * width + h * width + w;
reshapedArray[index] =
(float32Array[c * height * width + h * width + w] - 127.5) /
127.5;
}
}
}
return reshapedArray.reshape([1, 112, 112, 3]);
}
Recognition recognize(img.Image image, Rect location) {
var input = imageToArray(image);
// Prepare output for the model
List output = List.filled(1 * 192, 0).reshape([1, 192]);
final runs = DateTime.now().millisecondsSinceEpoch;
interpreter.run(input, output);
final run = DateTime.now().millisecondsSinceEpoch - runs;
print('Time to run inference: $run ms$output');
List<double> outputArray = output.first.cast<double>();
// Find the nearest registered face
Pair pair = findNearest(outputArray);
print("distance= ${pair.distance}");
// Create Recognition object
Recognition recognition =
Recognition(pair.name, location, outputArray, pair.distance);
// Save the recognition result
if (pair.name != "Unknown") {
saveRecognition(pair.name);
} else {
print("No match found. Recognition skipped.");
}
return recognition;
}
void saveRecognition(String name) async {
final timestamp = Timestamp.now();
await FirebaseFirestore.instance.collection('recognition_history').add({
'name': name,
'timestamp': timestamp,
});
print('Recognition saved for $name at $timestamp');
}
Pair findNearest(List<double> emb) {
Pair pair = Pair("Unknown", -5);
for (MapEntry<String, Recognition> item in registered.entries) {
final String name = item.key;
List<double> knownEmb = item.value.embeddings;
double distance = 0;
for (int i = 0; i < emb.length; i++) {
double diff = emb[i] - knownEmb[i];
distance += diff * diff;
}
distance = sqrt(distance);
if (pair.distance == -5 || distance < pair.distance) {
pair.distance = distance;
pair.name = name;
}
}
return pair;
}
void close() {
interpreter.close();
}
}
class Pair {
String name;
double distance;
Pair(this.name, this.distance);
}
